Li Keqiang leaves one wondering the 'could-haves' in Chinese semiconductor development

Misha Lu, DIGITIMES Asia, Taipei 0


"This might be one of the most advanced chips in the world," an official from the Belgium-based Interuniversity Microelectronics Centre (Imec) reportedly informed then-Chinese Premier Li Keqiang as he inspected a wafer during his 2018 visit to the semiconductor research institute. Three years before the visit, in June 2015, both Li and then-Belgian Prime Minister Charles Michel partook in a signing ceremony that saw Semiconductor Manufacturing International Corporation (SMIC), Huawei, Imec and Qualcomm co-investing to bring SMIC's 14nm process into volume production by 2020.

Only one year later in 2019, Huawei was added to the Entity List by the US, followed by full-blown US restrictions against China's semiconductor development at 16/14nm and below in October 2022.

With the recent passing of Li Keqiang, the term "Likonomics" was given a short injection of life as commentators search for Li's legacy in a ten-year tenure that saw the former premier gradually marginalized. Supposedly underpinned by the three pillars of "no stimulus", "deleveraging" and "structural reform", the term originally coined by Barclays analysts in the early days of Li's premiership quickly became void as it became hard to assess which aspects of Chinese economic and industrial policies are the brainchild of Li Keqiang. In its attempt to capture the technocrat's supply-side structural reform policy, it also neglected the emphasis to upgrade China's industry thorough a better allocation of investment capital.

In a turbulent 10-year period that saw the Chinese economy surpassing the US in purchasing power parity and its technology industry rapidly catching up under the benefits of globalization, all the while slowing to single-digit growth and running up against decoupling efforts and a trade war - perhaps, the legacy of a figure once widely expected to further Deng Xiaoping's reform and opening up policy is only endless "could-haves": would the course of China's development turn out any different, especially with its semiconductor industry very much at the heart of geopolitical competition?

As someone who prefers the "Hide your shine and bide your time" approach inherited from Deng and once openly objected to "leapfrogging" in technology development, preferring instead to overcome the "chokepoints" faced by Chinese technology industry via strengthening basic R&D - Li's passing left much space for imagination. Especially as leapfrogging has come to set the tone of Chinese technology development, despite what SMIC chief executive Liang Mong-Song famously remarked in 2021: there's no leapfrogging in chip manufacturing.

"Made in China 2025" a key pillar of structural reform

When examining China's semiconductor development during Li's premiership between 2013 and 2023 , "Made in China 2025" invariably came to the fore. Announced in 2015, the strategic roadmap, at its height, also embodied the Chinese tech industry's global aspiration.

Envisioning the country to become a leading manufacturing power between 2025 and 2030, Made in China 2025 is generally regarded as a pillar of Li's supply-side structural reform aiming to realign the mismatched supply and demand within a slowing Chinese economy long fueled by expansionary fiscal policy and burdened by overcapacity, over-stockpiling, excess liquidity and a real estate bubble, exacerbated by the CNY4 trillion stimulus package to tackle the 2008 financial crisis. However, the international attention drawn towards the project, thanks to aggressive government propaganda, was also partly to blame for the ongoing chip war.

Though Beijing has become low-key over the strategy, it still serves as a good reference to measure China's progress. Despite some of the lofty targets set - some with vague criteria - there still remain relatively clear benchmarks.

As indicated on the Made in China 2025 technology roadmap released by the government's National Manufacturing Power Construction Strategy Advisory Committee in 2015, a part of the plan aims to boost the Chinese semiconductor industry's total output to US$48.3- 85.1 billion, reaching 14.7% - 21.3% of the estimated global market size of US$328-400 billion by 2020, before elevating the total output to US$85.1 - 183.7 billion by 2030, reaching 21.3-34.2% of global market share. DIGITIMES Research's latest figures show that China's semiconductor industrial output was US$77 billion in 2022, accounting for 13% of the global market US$575 billion in size.

For the IC design sector, the goal was to reach US$40 billion worth of output, accounting for 25% of the global market by 2020, before reaching US$60 billion and 35% global market share by 2030. Figures from DIGITIMES Research show that the output of Chinese fabless companies only accounted for 15% of the global fabless market in 2022. If the output from IDMs is included, the share would be 9%. The target to reach 14nm design capability before 2020, however, is reached.

For the packaging sector, the goal was to reach US$10 billion worth of output, accounting for 35% of the global market by 2020, before reaching US$20 billion and 45% of global market share by 2030. DIGITIMES Research figures show that the output of the Chinese OSAT industry was US$10 billion in 2022, representing 24% of the global OSAT market. The goal to have advanced 3D packaging capability before before 2020, meanwhile, is not yet met, though JCET and Tongfu Microelectronics can do 2.5D packaging.

When it comes to chip manufacturing, the roadmap seeks to reach 20-14nm manufacturing capability by 2025, which is already achieved. It also aims for a monthly installed capacity of one million 300mm wafers by 2025. Based on SEMI's 300mm Fab Outlook to 2026, Chinese 300mm wafer capacity was 1,540,000 in 2022. Assuming the figure includes foreign-invested capacity, if TSMC, UMC, Powerchip, Samsung and SK Hyix's capacities in China are deduced, the figure would be approximately 860,000 wafers per month.

Certainly, only a glimpse into the semiconductor aspect of Made in China 2025 is offered here, omitting other variables of Chinese chip industry development such as the well-known issues of corruption, misallocation of capital and the potential problem with overcapacity. Ultimately, only time will reveal the true extent of Li's influence on China's semiconductor development course.